Background: Osteochondritis dissecans (OCD) healing can be unpredictably slow, incomplete, or absent after surgical treatment. This frustrates patients, families, and the medical team. We aimed to develop an algorithm to predict the speed of OCD radiographic ossification based on patient demographic, physical, surgical, and imaging data. Methods: We studied a prospective cohort of patients with knee OCD lesions in a multicenter database. We included patients who were diagnosed with knee OCD lesions of the lateral or medial femoral condyle and were treated operatively at a single center. We collected patient information from medical records and imaging studies. Radiographic healing was defined on the basis of the percentage of the original OCD lesion that had a normal bone density (ossification) compared with the surrounding condyle, rated on a continuous scale from 0 to 100. An OCD lesion that achieved ≥90% of the normal surrounding bone density at 6 months following surgery was defined as fast healing. Follow-up was conducted with radiographs only. Multivariable logistic regression and receiver operating characteristic (ROC) curve analyses were performed. Results: This study included 79 OCD lesions in 72 individuals. The mean patient age was 13.79 ± 2.71 years (range, 8.56 to 22.98 years), and 45 (62.5%) of the 72 patients were male. A total of 56 patients (77.8%) were White, and 69 patients (95.8%) were of non-Hispanic ethnicity. In all, 23 lesions (29.1%) fit the of fast healing. A multivariable regression analysis revealed that high preoperative bone density within the OCD lesion (p < 0.001) was the only feature that had an association with fast healing. A preoperative lesion density rating of ≥70% predicted fast ossification with a sensitivity of 87% and a specificity of 66.1%. Conclusions: A preoperative OCD density rating that was ≥70% of that of the normal surrounding bone was found to be the only independent predictor of fast ossification following surgery. Level of Evidence: Prognostic Level II . See Instructions for Authors for a complete description of levels of evidence.
Segal et al. (Thu,) studied this question.